Calculate the probability of reversal - page 5
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
I support Bas with his advice - you need to move into options. The Black-Scholes model should obviously work on your data.
You didn't even understand what the question was about and what my advice was about) why show your ignorance so clearly?)
Is every bell-shaped function a density of a normal distribution? What prevents you, for example, from seeing in your figure the density of the beta distribution?
By the way, Alexey, and Vladimir, can you give us a hint? Suppose we want to approximate some data by a normal distribution.
Tails and middle of distribution must have the same weight in approximation as I suppose?
Then it is better to approximate in logarithmic coordinates?
For in linear coordinates the absolute error in the tails will be orders of magnitude smaller than in the middle, and thus will poorly participate in the approximation.
Or the second option - to take the quotient, not the square of the difference, as the error? But I will not be able to derive such formulas.
You didn't even understand what the question was about and what my advice was about) why show your ignorance so clearly?)
By the way, Alexey, and Vladimir, can you give us a hint? Suppose we want to approximate some data by a normal distribution.
Tails and middle of distribution should have the same weight in approximation, as I suppose?
Then it is better to approximate in logarithmic coordinates?
For in linear coordinates the absolute error in the tails will be orders of magnitude smaller than in the middle, and thus will poorly participate in the approximation.
Or the second option - to take the quotient, not the square of the difference, as the error? But I can't derive such formulas.
First of all it would be necessary to check normality of the sample, at least "by eye".
Then you should find "applied statistics" by Kobzar and look there the second chapter)
...
Would it then be better to approximate in logarithmic coordinates?
...In general your question (questions?) is put too general, abstractly. The only thing I can say with certainty - if approximation is done with minimization of deviations in logarithmic coordinates, it will be minimum relative error. This applies to approximation by anything: polynomials, trigonometric functions, splines, rational fractions, wavelets... I have not heard of approximations with typical probability distributions.
To begin with, it is worth checking that the sample is normal, at least "by eye".
"This means to construct a quantile-quantile(or probability-probability) plot for the sample and the normal distribution and make sure that it is well approximated by a straight line.
By the way, Alexey, and Vladimir, tell me. Suppose we want to approximate some data by a normal distribution
..
Here, for example, is the data that everyone is interested in right now https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6. Just "some" data.
Why would they need to be approximated by a normal or other distribution? The exponent was still interesting in the beginning, according to the chain reaction kinetics equation of propagation.
Why would they need to be approximated by a normal or other distribution? The exponent was interesting in the beginning, according to the chain reaction propagation kinetics equation.
What is meant is not a time series, but a histogram close to normal.
https://www.mql5.com/ru/articles/396
python -https://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.gaussian_kde.html
https://www.mql5.com/ru/articles/396
So it's smoothing.